seo-page
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from external URLs and local files, which presents a surface for potential injection of instructions embedded in web content.
- Ingestion points: Content fetched from URLs via WebFetch or read from local files (SKILL.md).
- Boundary markers: None specified to differentiate user instructions from data content.
- Capability inventory: Executes local analysis scripts
scripts/parse-meta.ts,scripts/validate-schema.ts, andscripts/geo-score.ts(SKILL.md). - Sanitization: Not explicitly mentioned in the workflow.
- [COMMAND_EXECUTION]: The workflow involves running local TypeScript scripts to process inputs.
- Evidence: Mentions running
scripts/parse-meta.ts,scripts/validate-schema.ts, andscripts/geo-score.tswith input arguments (SKILL.md).
Audit Metadata